101 results about "Feature mining" patented technology

The present application relates to a kind of based on multi-source industrial data's car body welding quality intelligent perception, prediction and decision-making method and system, belong to the technical field of intelligent manufacturing and industrial data analysis.The present application obtains multi-source high-frequency synchronous original signal by sensor array and robot bus;Perform space-time alignment and multidimensional feature mining;Quality perception model is constructed based on multi-path parallel network and cross-modal attention;Quality evolution trend and electrode cap life are evaluated using time series prediction model;Integrate expert rules and reinforcement learning to make process compensation decision.The present application realizes the accurate perception of welding quality, forward warning and closed-loop real-time compensation, significantly improves the perception accuracy, production robustness and prolongs the service life of electrode.

An optical remote sensing image processing system and training method, relating to the fields of computer vision and image processing technology, alleviates the difficulties in balancing high accuracy and high efficiency in saliency enhancement in existing optical remote sensing image processing technologies, which suffer from high computational costs, incomplete structures, and unclear boundaries. The optical remote sensing image processing system includes a basic feature extraction network module for extracting basic features from the optical remote sensing image to be enhanced; a multi-directional feature mining and aggregation module for obtaining corresponding feature map sequences based on the basic feature map sequences; a cross-scale edge information fusion module for fusing the feature map sequences through cross-scale edge scanning; and a saliency enhancement module for obtaining a refined saliency map. This invention is applicable to remote sensing and UAV image analysis. It can quickly identify salient areas on the Earth's surface, such as buildings, ships, and disaster areas.

The embodiment of the application provides a network traffic map data processing method and system applied to digital services, determines a target network traffic map sequence and a network behavior mining node sequence corresponding to each target network traffic map in the target network traffic map sequence, obtains an initial node change instruction corresponding to the target network traffic map sequence, determines a changed network traffic map sequence corresponding to the target network traffic map sequence according to the initial node change instruction, obtains a first model scheduling distinguishing parameter corresponding to the changed network traffic map sequence, obtains a second model scheduling distinguishing parameter corresponding to the changed network traffic map sequence, determines a target change instruction according to the first model scheduling distinguishing parameter and the second model scheduling distinguishing parameter corresponding to each initial node change instruction, and changes the network behavior mining node in the target network traffic map sequence according to the target change instruction, thereby improving the feature mining reliability of the network traffic map.

This invention discloses a short-term intelligent numerical weather prediction method and system based on environmental perception, belonging to the field of meteorological management technology. The system includes a data acquisition and fusion module, an intelligent zoning and feature mining module, a short-term forecasting and risk assessment module, and a visualization and focusing module. The data acquisition and fusion module collects and fuses environmental data from meteorological radar and weather stations, as well as GIS maps. The intelligent zoning and feature mining module divides the data into grids based on the GIS map, analyzes the rainfall synergy between grids based on historical environmental data to construct analysis groups, and selects and constructs an impact set for each analysis group. The short-term forecasting and risk assessment module fits near-term precipitation intensity relationships and short-term precipitation intensity relationships based on the impact sets of each analysis group. The visualization and focusing module calculates and outputs the near-term and short-term predicted precipitation intensities for each analysis group based on the relationships, identifies key areas of concern using topographic and hydrological models, and enhances visualization.

This invention relates to the interdisciplinary field of network security and artificial intelligence, and discloses a method for anomaly mining and sample generation in encrypted traffic. The method includes: data collection and preprocessing to construct a dataset of normal encrypted traffic under multiple scenarios and a dataset of a small number of real malicious samples; deep behavioral modeling of normal encrypted traffic under multiple scenarios based on a large model to construct a baseline security behavior knowledge base; fine-grained comparison of the features of the encrypted traffic to be detected with normal traffic based on an improved contrastive learning algorithm to mine hidden anomaly features; generating encrypted malicious samples based on a controllable generation architecture of the large model, combining hidden anomaly features and known attack patterns; optimizing the encrypted traffic anomaly detection model using the generated encrypted malicious samples, and feeding the optimized encrypted traffic anomaly detection model back to the anomaly feature mining stage to form a closed-loop iterative optimization system. Using this invention, the accuracy, generalization, and real-time performance of encrypted traffic anomaly detection can be improved.

This invention discloses a lightweight image restoration method, relating to the fields of computer vision and image processing technology, comprising: S1, input preprocessing: inputting a damaged image and a size-matched binary mask, concatenating the two along the channel dimension to obtain an input tensor; S2, encoder feature extraction: configuring an encoder network composed of multiple downsampling blocks to extract multi-scale features; the downsampling block includes an LSConv module with improved LSNet convolution, convolutional layers, normalization layers, and activation functions. The LSConv module, with its separable convolutional structure based on multi-branch depth, simultaneously captures global structure and local detail features. Feature correction and fusion are completed through feature concatenation and channel adjustment combined with the SE attention mechanism. This invention innovatively designs a novel LSConv feature extraction module, employing a multi-branch parallel structure combined with multi-scale feature mining, breaking through the limitations of traditional single convolution, simultaneously capturing global and local features, and enhancing feature expression through channel optimization, thereby improving image restoration capabilities.

This invention belongs to the field of acute lymphoblastic leukemia (ALL) diagnosis and treatment technology, and discloses an ALL scoring system. The feature library construction module establishes an associated feature system based on subtype molecular biological differences, uses LASSO regression and subtype stratified analysis to screen core features and determine weights, and excludes redundant cross-subtype information. The feature mining module targets the entire treatment cycle, dividing the collection nodes into induction, consolidation, and maintenance phases, capturing temporal correlation features such as the rate of MRD decline and gene expression change rate, breaking the limitation of relying solely on static data at the time of diagnosis. It not only fits the individual characteristics of patients with different subtypes, but also reflects the dynamic changes in the treatment process, improving the accuracy of prognostic assessment. A real-time corrected score is obtained by calculating the baseline score and the temporal feature adjustment score. Patient data is updated weekly through the electronic medical record interface, and appropriate solutions are output for different situations such as low-risk, intermediate-risk, and very high-risk.

The application relates to a used commodity feature mining method and system, wherein the method comprises the following steps: reading a commodity quality inspection report to obtain text information; extracting one or more pending commodity features from the text information; adding the pending commodity features to a first information feature set of a function model to form a second information feature set; taking the first information feature set and the second information feature set as inputs of the function model to obtain a first prediction evaluation value and a second prediction evaluation value; and comparing the first prediction evaluation value and the second prediction evaluation value; and in response to the second prediction evaluation value being greater than the first prediction evaluation value, confirming that the pending commodity features are available. The application extracts available and effective commodity features from a commodity quality inspection report, thereby refining the description of the commodity, enabling downstream storage such as search, recommendation and other services to better distinguish commodities and provide more accurate recall commodities.

The application relates to the technical field of artificial intelligence, in particular to a model optimization method and device based on an intelligent agent. A model optimization instruction is understood by an intention understanding intelligent agent to obtain an intelligent agent scheduling task instruction, a target intelligent agent is called in sequence according to the scheduling sequence of the intelligent agent in the intelligent agent scheduling task instruction, data to be processed is sent to the target intelligent agent to complete output processing until a model parameter adjustment suggestion output by the last intelligent agent in the scheduling sequence is obtained, and parameters in a to-be-optimized model are adjusted based on the model parameter adjustment suggestion. Since the callable intelligent agents include a data query intelligent agent, a code execution intelligent agent, a feature mining intelligent agent, a model evaluation intelligent agent and a model optimization intelligent agent, when the model is optimized, automatic data query, feature analysis, model evaluation and optimization operations can be performed based on the callable intelligent agents, and the efficiency of model optimization is improved.

A point cloud classification and segmentation method based on dual-path feature fusion, comprising the following steps: 1) preprocessing the point cloud classification and segmentation dataset; 2) constructing a dual-branch feature extraction structure, wherein branch one adopts a multi-layer perceptron and a hybrid pooling to realize global structure feature extraction, and branch two mines point cloud local neighborhood geometric correlation features through a local cross-attention mechanism; 3) building a dual-path feature fusion network, training the model by using a training set, and completing effective fusion of global and local features; and 4) loading the trained network parameters, performing end-to-end inference on a test set, and outputting point cloud classification results and point-by-point segmentation results. Through the dual-path differentiated feature extraction and fusion design, the global structure representation and local detail perception are taken into account, so as to solve the technical problems that in the present point cloud classification and segmentation task, excessive focus is placed on global feature extraction, local feature mining is insufficient, and the classification accuracy and segmentation granularity are insufficient.

The application discloses a strategy for extracting fault features of electronic circuit systems, named as SSEST strategy, which is used for perceiving global information and paying attention to notable local information, and mining important local information means realizing expression of intermittent fault features of electronic circuit systems, specifically, first, S transformation is performed on a circuit output time sequence signal to acquire time-frequency domain features, then a squeeze and excitation network attention module is used to distribute channel weights, subsequently, input into a Swin Transformer framework, and pay attention to local information related to faults from global signals, and deep mining is performed on fault features, and two electronic circuits are taken as experimental circuits, the proposed diagnostic strategy realizes rapid and high-precision diagnosis, and shows that the proposed multiple attention mechanism is efficient for feature mining of intermittent faults of electronic circuit systems.

The application discloses a photovoltaic cluster power prediction method based on similar station related feature mining, and relates to the technical field of photovoltaic power prediction.The method comprises the following steps: dividing a to-be-predicted day into a sunny day prediction day and a non-sunny day prediction day; for the sunny day prediction day, directly using a pre-trained time convolution network model for prediction; for the non-sunny day prediction day, constructing a historical theoretical power difference value of a target station, simultaneously screening similar stations according to power generation and weather characteristics, determining historical similar weather characteristics, inputting into a time convolution neural network, predicting a power difference value, and calculating a non-sunny day power prediction result; and performing power prediction on all photovoltaic stations in a photovoltaic cluster to obtain a predicted power of the photovoltaic cluster. By analyzing weather characteristics of adjacent stations and calculating a historical theoretical power difference value of a target station, a time convolution network model is applied to capture long-term dependence in a sequence, and the prediction accuracy and prediction reliability are improved.

The technical scheme of the present application discloses a kidney transplantation single disease follow-up database system based on a hospital intranet environment, comprising a data integration layer, an intelligent analysis layer and a doctor-patient linkage layer. Through the cooperative operation of the three-layer technical architecture designed by the present application, the three core problems of the prior art are comprehensively solved, and the following technical effects are achieved: at the data integration level, automatic and real-time integration of multi-port medical data of kidney transplantation patients is realized, the data integration efficiency is improved by 90%, the accuracy rate reaches 99.5%, and the traceability and privacy security are significantly improved; at the data analysis level, the AI-driven analysis model realizes automatic generation of patient whole-cycle health reports, large-sample group feature mining and real-time abnormality early warning, and the doctor data analysis workload is reduced by 80%, and the abnormal data response time is shortened to the minute level.

This invention discloses a lightweight method and system for identifying smoke areas in photovoltaic (PV) plants based on an improved YOLO framework. It relates to the field of image recognition technology and includes: multi-point monitoring feature mining based on historical smoke detection logs of the PV plant; real-time monitoring of a multi-sensor array controlled by a monitoring attention-configured graph network to obtain visual and thermal image streams of the plant area; introducing an improved YOLO multi-level learning architecture for visual and thermal smoke probability detection; interference tracing correction and time-dependent aggregation of the first and second smoke detection graph networks; and performing smoke area identification based on the reliable smoke detection graph network to obtain a smoke area identification graph network. This invention solves the technical problems of low accuracy, poor real-time performance, and insufficient reliability in smoke area identification in complex and vast PV plant environments, achieving accurate, reliable, and adaptively focused real-time smoke identification in PV plant areas.

The application discloses a diaphragm electromyographic stimulation waveform optimization method and device based on energy control, and relates to the technical fields of energy measurement and control and artificial intelligence. The method comprises the following steps: collecting multi-source data reflecting an electric stimulation process, preprocessing the multi-source data, and constructing a training data set based on the preprocessed multi-source data; constructing an intelligent evaluation model integrating time sequence feature mining and double-task prediction functions, and training the intelligent evaluation model in stages according to the training data set, so that the model has the joint prediction ability of the neural recruitment efficiency and the subjective comfort degree of different stimulation waveforms and parameter combinations; based on the joint prediction ability of the intelligent evaluation model, automatically optimizing the key hyperparameters of the model by using an elite base butterfly optimization algorithm; and taking the dynamic constancy of a preset target stimulation energy as a constraint condition, and deciding the optimal stimulation waveform and parameters according to the trained and hyperparameter-optimized intelligent evaluation model, so that the dual improvement of stimulation efficacy and safety is realized.

The application relates to the technical field of energetic material design and performance prediction, and relates to an energetic material multitask property prediction method. Standardized data sets are constructed by collecting energetic material sample data and performing molecular graph data enhancement processing on a training subset; molecular SMILES strings are respectively parsed into molecular topological graphs and word sequences, three-dimensional graph neural network is used to encode atomic space position information to extract molecular local topological features, a Transformer encoder is used to extract molecular global long-range correlation features, and a dynamic weighting mechanism is used to realize adaptive fusion of the two types of features, so that the depth and comprehensiveness of feature mining are improved; a multitask prediction model with a shared encoding layer, multiple independent decoding layers and a task decoupling constraint mechanism is built, an end-to-end training is completed by using a multitask loss function, the problems that existing multitask methods are difficult to balance the internal conflicts between energy performance and safety performance and are difficult to simultaneously predict multiple key indicators are solved, and the simultaneous prediction of multiple performance indicators is realized.

This invention relates to a deep learning-based system and method for predicting hematoma expansion in spontaneous intracerebral hemorrhage. The system includes a data acquisition module for acquiring the patient's baseline head CT image data, corresponding clinical information, and blood index data; a quantization feature calculation module for performing three-dimensional hematoma segmentation on the CT image data to obtain a three-dimensional hematoma mask; a multi-dimensional feature extraction module, including a three-dimensional morphological feature extraction sub-network and a temporal feature extraction sub-network; a feature fusion module for acquiring high-dimensional features; and a risk prediction module for outputting the risk probability of hematoma expansion based on the high-dimensional features and generating a corresponding visualized risk factor weight map. This system deeply integrates data-driven three-dimensional morphological depth features and temporal features with knowledge-driven surface regularity index and density variation coefficient quantization indicators, improving feature mining capabilities while ensuring sensitivity to key pathological morphologies, thus enhancing the robustness and accuracy of prediction.

This invention discloses an AI-based smart library borrowing management system, belonging to the field of library borrowing management technology. The AI-based smart library borrowing management system includes: a data acquisition module for acquiring target user book borrowing information and the status information of books to be borrowed; a user feature mining module for extracting borrowing behavior feature sets based on book borrowing information and generating a borrowing credit risk assessment value; a book status analysis module for extracting book status feature sets based on book status information and generating a borrowing suitability assessment value; an association matching analysis module for generating a borrowing matching assessment value by interactively fusing two types of feature sets through an association matching model; and a comprehensive assessment module for generating a comprehensive borrowing assessment value by combining the above assessment values. This invention manages the borrowing of target users through the comprehensive borrowing assessment value in the borrowing management module, thereby achieving precise matching between user borrowing needs and book resources, and thus improving the accuracy of management decisions.

The application discloses a power distribution network topology identification method and system based on artificial intelligence and belongs to the technical field of power system automation. Through KNN missing value filling, photovoltaic output scene clustering and normalization preprocessing on the measurement data of the power distribution network, random forest is used to screen voltage characteristics to reduce the input dimension, and an RF-CNN-Attention hybrid model is constructed, wherein the CNN extracts local characteristics, the attention mechanism calculates attention weights by generating query, key and value matrices and gives different weights to feature maps, combines residual connection to enhance feature expression, and finally realizes topology label classification through a full connection layer and the like. Through multi-model fusion and fine feature processing, the application improves the feature mining capability and processing efficiency of the power distribution network topology identification.

This application discloses a method, apparatus, equipment, and storage medium for predicting the moisture content of tobacco shreds at the outlet, relating to the field of tobacco technology. The method addresses the challenge of accurately predicting the moisture content of tobacco shreds at the outlet due to the strong coupling of multiple variables, nonlinearity, and large time delays in the tobacco drying process. It also addresses the pain points of existing mechanistic models (insufficient generalization), traditional recurrent neural networks (RNNs) prone to gradient anomalies in long-sequence modeling, and the high computational cost and difficulty in adapting to the real-time and deployment requirements of mainstream attention models for industrial online soft measurement. This method constructs a standardized prediction model system adapted to the time-series characteristics of tobacco drying. Through full-process time-series data normalization and core feature mining, feature biases caused by noise and time delays are eliminated. Lightweight long-sequence modeling optimization balances long-range dynamic capture capability and computational cost, avoids gradient anomalies, breaks through quadratic complexity limitations, reduces resource consumption, and provides stable soft measurement data for outlet moisture content.

The small sample point cloud target classification method based on LSHW-Mamba network relates to the technical field of 3D point cloud processing and deep learning, and solves the problems that the existing technology focuses on feature mining in the Euclidean space domain, and under the condition of insufficient target samples, the feature representation ability of the target geometric structure is insufficient, and it is difficult to capture the key information for distinguishing similar objects. The present application explicitly analyzes the spectral response of the local geometric structure of the point cloud through the spherical harmonic function, combines the local spherical harmonic wavelet transform to realize the geometric feature extraction of the point cloud, simultaneously introduces the geometric gating mechanism to optimize the feature propagation efficiency of the Mamba model, models the target features in parallel with the Transformer attention branch, considers the relevance of the local and global features of the target, and improves the discrimination ability of the network model for the target under the small sample scene. The present application effectively solves the core problems of insufficient feature discrimination and decreased classification accuracy under the small sample condition.

This invention relates to a lithium battery state of health estimation method based on meta-transfer learning, belonging to the interdisciplinary field of battery management systems and signal processing. This method enhances feature mining capabilities through a cascaded structure of a 1-D CNN and a Transformer encoder. Simultaneously, it integrates meta-learning and transfer learning mechanisms. Meta-learning optimizes model initialization parameters across multiple source tasks, reducing the dependence on the similarity between the source and target domain distributions; transfer learning, through a pre-training-fine-tuning paradigm, reduces the requirements for the number and diversity of source tasks. This joint mechanism overcomes the limitations of traditional single learning methods in cross-domain generalization, maintaining high SOH estimation accuracy even when there are significant differences in distributions between domains.

The application discloses a resting EEG psychological stress prediction method fusing bidirectional decoupling and potential feature mining, which is applied to a resting EEG psychological stress prediction system fusing bidirectional decoupling and potential feature mining. The prediction system is obtained by training according to historical electroencephalogram (EEG) data and corresponding psychological stress labels. The prediction method comprises the following steps: preprocessing collected original EEG data, and dividing the preprocessed continuous EEG data into a plurality of slices according to a sliding time window, wherein each slice comprises a task-state EEG signal and a resting-state EEG signal; adding random noise to the task-state EEG signal, so that the noisy task-state EEG signal is randomly distributed; introducing a guiding mechanism of the task-state electroencephalogram to the resting-state potential feature, so as to improve the feature separability in a weak feature scenario; meanwhile, effectively separating individual differences and depression-related features by using a bidirectional decoupling structure, extracting the time-space features of the EEG by using a wavelet-Riemann technology, and improving the psychological stress prediction efficiency and objectivity.

This invention belongs to the field of industrial intelligent monitoring and structural deformation sensing technology, and relates to a method and system for intelligent monitoring of crankshaft deformation. The method includes: acquiring original strain signals and rotational speed pulse signals; extracting a reference envelope surface based on the rotational speed pulses, fusing temperature and centrifugal parameters to construct a compensation baseline and subtracting it to generate a dynamic bending signal; converting the signal into offset data and calculating the bend difference data; applying cubic spline interpolation to generate a continuous deflection curve and extracting representative deformation values ​​within the window to calculate the differential slope; performing extrapolation calculations when the slope continuously increases and exceeds the limit, outputting the remaining safety margin and triggering an early warning. The technical solution of this application achieves accurate restoration and trend tracking of real bending characteristics through multi-dimensional environmental baseline compensation and spatiotemporal feature mining, which can quantify the remaining safe life and capture fatigue damage in advance, meeting the needs of accurate perception and predictive maintenance of crankshaft operating status under complex working conditions.

The application discloses a cancer molecular marker prediction method based on multi-omics and intelligent feature mining, and belongs to the technical field of bioinformatics and artificial intelligence; the method comprises the following steps: collecting multi-omics data of cancer samples, including genomic, transcriptomic, epigenetic, proteomic and metabolomic data; performing quality control and pretreatment on the data; performing intelligent feature mining by adopting difference analysis, dimension reduction algorithm and feature fusion network; integrating biological networks to construct a multi-view graph neural network prediction model; identifying key molecular markers and performing verification; the application can effectively integrate multi-level omics data, mine molecular markers with biological significance, and establish a high-precision and high-explainability prediction model; the application provides an effective technical means for early diagnosis, molecular typing, prognosis evaluation and individualized treatment of cancer, and has important scientific significance and clinical application value.

This invention discloses a fault diagnosis method for crane maintenance management, relating to the field of crane equipment operation and maintenance technology. It includes collecting a multi-source sensor data sequence consisting of vibration spectrum of the hoisting mechanism, wire rope tension fluctuations, and temperature changes at hinge points within the crane's operating cycle. The multi-source sensor data sequence is time-synchronized and outlier-removed to generate a standardized synchronization data matrix. The synchronization data matrix is ​​then fed into a hierarchical fault feature extraction process, outputting degradation feature vectors for key components and anomaly pattern identifiers. The remaining effective working range length of the components is calculated using the degradation feature vectors, and the fault initiation time point is located using the anomaly pattern identifier. These two types of information are merged and encoded to generate a maintenance trigger signal, which is then pushed to the crane control terminal display interface. This method achieves standardized organization and hierarchical feature mining of multi-source operating data, quantifies the service status of components, and identifies the fault initiation time, providing a standardized status basis for crane maintenance management.

This invention discloses a method for constructing a time-series prediction model for the water environment based on dynamic feature mining of data, belonging to the field of environmental engineering technology. It solves the problems of high complexity and low efficiency in existing methods for dynamic feature mining of water environment data. The invention identifies target variables for a target watershed, collects historical data of the target variables and related variables, extracts features from the historical data, and establishes an initial input dataset. It uses grey relational analysis to analyze the correlation between the features of the target variables and the features of related variables, filtering variables whose correlation with the target variables exceeds a threshold. It uses SSA to optimize the parameters of STL, and uses the optimized STL to decompose the historical data of the target variables. Finally, it establishes an LSTM model, trains the LSTM model using the decomposed data and variables whose correlation exceeds the threshold, and obtains the SSA-LSTM model. This invention is applicable to water environment prediction and dynamic feature mining.

The present application relates to a kind of industrial solid waste treatment equipment remote diagnosis method, comprising: by the distributed collection equipment of multiple types of sensors different parts of multimodal real-time working condition signal, in combination with geographical position and working condition label, establish position annotation, working condition mapping multi-source original data set;Based on statistical, time domain and frequency domain feature mining, generate high-dimensional feature node that fuses space-time information;Accordingly, construct space-time evolution graph network, and introduce dynamic graph neural network to learn the dynamic association between nodes and working condition evolution law, realize the high-precision identification of complex anomaly and precursor track extraction, finally generate traceable multi-label anomaly early warning.

This invention discloses a data fusion and knowledge graph construction system and method for intelligent laboratories. The system includes a data acquisition layer, a data parsing layer, a feature recognition layer, a construction layer, and a knowledge update layer. The invention achieves this by simultaneously acquiring IoT sensor data and experimental operator verbal data through the data acquisition layer, and performing time-axis-based voice-sensor data modal alignment through the data parsing layer. This deeply integrates unstructured operational descriptions with structured physical parameters into standardized experimental records. The feature recognition layer automatically triggers a failure feature mining process when experimental results fail to meet standards, encapsulating operational deviations and material / environmental disturbances into negative feature triples and storing them in the knowledge graph. Furthermore, the knowledge update layer utilizes graph representation learning technology for cross-batch similarity analysis, dynamically identifying the impact of latent variables outside standard operating procedures on experimental results. This enables dynamic adjustment of the knowledge graph structure and subsequent experimental risk warnings.